Currently, on stm32f cortex-m0 systems, interrupts are always being
directed to the RO vector table. This can cause strange problems when
running RW software because it is still calling IRQ handlers in the RO
code.
Unfortunately, on cortex-m0 the ability to specify the vector table
location in flash (VTOR register) is optional, and stm32f0 parts do not
have it. Instead, in order to run RW IRQ handlers, at init time, this
CL copies the vector table from flash to the base address of SRAM
(0x20000000), and then selects SRAM to be mapped to 0x00000000 where
the core looks to find the vector table.
BUG=none
BRANCH=none
TEST=Tested on zinger.
- Verified that vector table is copied to SRAM by printing out 48 words
from SRAM base address 0x20000000 in main() and verifying that it matches
the vector table in flash in the disassembly.
- Verified the vector table at SRAM 0x20000000 points to the RW handlers
when in RW and the RO handlers when in RO.
- Also printed out PC in one IRQ handler and verified it was in the
appropriate section of code.
Also, ran on samus_pd and did a sysjump RW to make sure at least one other
system works.
Change-Id: I22aff1b5e0de9b23fd3324f0cbe4f6c45a81967e
Signed-off-by: Alec Berg <alecaberg@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/210063
Reviewed-by: Vic Yang <victoryang@chromium.org>
Reviewed-by: Vincent Palatin <vpalatin@chromium.org>
Import code to do 64-bit multiplication on Cortex-M0 core without SMULL
instruction.
Signed-off-by: Vincent Palatin <vpalatin@chromium.org>
BRANCH=none
BUG=none
TEST=make buildall
add a 64-bit multiplication and see it compiled properly.
verify in .map files that the code is discarded for cortex-M0 based
platforms not using the 64-bit multiplication.
Change-Id: I0a91b3502f4bee4bb79b193fe0854e56a7d498f7
Reviewed-on: https://chromium-review.googlesource.com/207132
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Tested-by: Vincent Palatin <vpalatin@chromium.org>
Commit-Queue: Vincent Palatin <vpalatin@chromium.org>
On panic, reboot properly the CPU rather than just jumping to the reset
vector as that might lead to some incorrect initializations.
Properly plug the div by 0 to the panic handling.
Add a small trace if the debug output is activated.
Signed-off-by: Vincent Palatin <vpalatin@chromium.org>
BRANCH=none
BUG=chrome-os-partner:29840
TEST=add adhoc code triggering a data abort and see the firmware
printing a trace, then rebooting immediatly in a working state.
Change-Id: I1d5a98d9113c8ae08e05588a40f941d1ed22cebe
Reviewed-on: https://chromium-review.googlesource.com/206268
Reviewed-by: Vic Yang <victoryang@chromium.org>
Tested-by: Vincent Palatin <vpalatin@chromium.org>
Commit-Queue: Vincent Palatin <vpalatin@chromium.org>
Enough USB support to be able to enumerate the device and use bulk or
interrupt endpoints.
Signed-off-by: Vincent Palatin <vpalatin@chromium.org>
BRANCH=none
BUG=chrome-os-partner:28295
TEST=with the following USB console CL, connect a Fruitpie through USB
and use its console over USB.
Change-Id: I37b7f3b6a754cb82ab5f940ea20122d2e16b3b5b
Reviewed-on: https://chromium-review.googlesource.com/193983
Tested-by: Vincent Palatin <vpalatin@chromium.org>
Reviewed-by: Todd Broch <tbroch@chromium.org>
Commit-Queue: Vincent Palatin <vpalatin@chromium.org>
Our code base contains a lot of debug messages in this pattern:
CPRINTF("[%T xxx]\n") or ccprintf("[%T xxx]\n")
The strings are taking up spaces in the EC binaries, so let's refactor
this by adding cprints() and ccprints().
cprints() is just like cprintf(), except that it adds the brackets
and the timestamp. ccprints() is equivalent to cprints(CC_CONSOLE, ...)
This saves us hundreds of bytes in EC binaries.
BUG=chromium:374575
TEST=Build and check flash size
BRANCH=None
Change-Id: Ifafe8dc1b80e698b28ed42b70518c7917b49ee51
Signed-off-by: Vic Yang <victoryang@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/200490
Reviewed-by: Randall Spangler <rspangler@chromium.org>
This is another patch to fix the bug which causes a HardFault exception
at the "svc" instruction in __wait_evt().
The HardFault is due to a priority escalation problem in which "svc"
is called when the PRIMASK is high, meaning interrupts are disabled.
The issue was that an interrupt can occur just before the "svc"
instruction, and when an interrupt fires that performs a context switch,
the IRQ handler disables interrupts setting the PRIMASK reg high.
The arm v6 reference manual specifies that "PRIMASK unchanged on
exception exit". So, therefore, we must clear PRIMASK by running "cpsie"
before exiting IRQ handler.
BRANCH=none
BUG=chrome-os-partner:28296
TEST=
Reproduce the problem on a fruitpie by inserting dummy for loop in
__wait_evt() before "svc" call:
asm volatile("isb");
for (i = 0; i < 250; i++) ;
__schedule(1, resched);
Then, when running pd dev, the system gets the HardFault exception
within a few minutes because there is more time for an interrupt
to occur and disable interrupts right before call to "svc".
After applying this patch, the code has run for > 3 hours without
a HardFault.
Change-Id: Ic50252b09c40c7d76975ff7f16d799c9eae2bde6
Signed-off-by: Alec Berg <alecaberg@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/197839
Reviewed-by: Vincent Palatin <vpalatin@chromium.org>
From time to time (usually under heavy interrupt load), the runtime on
Cortex-M0 was panic'ing at the "svc" instruction with a HardFault
exception (inside the wait_event() function).
The issue was probably the following :
the wait_event() code is doing an atomic_read_clear() whose critical
section disables interrupts and re-enables them using "cpsie i",
then do __schedule() call which is essentially a "svc" instruction.
According to ARMv6-m reference manual :
"If execution of a CPS instruction:
increases the execution priority, the CPS execution serializes that
change to the instruction stream.
decreases the execution priority, the architecture guarantees only that
the new priority is visible to instructions executed after either executing
an ISB instruction, or performing an exception entry or exception
return."
So, when we are executing the "svc", PRIMASK.PM can still be seen as 1
(while it was set to 0 by "cpsie i") and in that case the software
interrupt is replaced by a HardFault.
Signed-off-by: Vincent Palatin <vpalatin@chromium.org>
BRANCH=none
BUG=chrome-os-partner:28296
TEST=run Firefly board under load for extended periods of time.
Change-Id: Ie355c36f06e6fe2fee5cca8998a469fa096badad
Reviewed-on: https://chromium-review.googlesource.com/196659
Commit-Queue: Vincent Palatin <vpalatin@chromium.org>
Tested-by: Vincent Palatin <vpalatin@chromium.org>
Reviewed-by: Alec Berg <alecaberg@chromium.org>
In order to achieve really tiny firmwares, make our runtime (tasks,
hooks, muxed timers, GPIO abstraction ...) optional.
Add 2 new build options for it : CONFIG_COMMON_RUNTIME and
CONFIG_COMMON_GPIO which are enabled by default, and ensure all the
source files are built according to the right configuration variable.
Signed-off-by: Vincent Palatin <vpalatin@chromium.org>
BRANCH=none
BUG=none
TEST=make buildall
build a minimal board with no runtime.
Change-Id: Icb621cbe0a75b3a320cb53c3267d6e578cd3c32f
Reviewed-on: https://chromium-review.googlesource.com/189403
Reviewed-by: Vic Yang <victoryang@chromium.org>
Commit-Queue: Vincent Palatin <vpalatin@chromium.org>
Tested-by: Vincent Palatin <vpalatin@chromium.org>
Put each functions in a separate section by using -ffunction-sections,
then discard the non-referenced ones in the linker with -gc-sections.
Force keeping a few special symbols by using the KEEP() linker
directive.
This modification is not saving a lot of spaces per se, but will enable
larger code pruning with future optional features.
Signed-off-by: Vincent Palatin <vpalatin@chromium.org>
BRANCH=none
BUG=none
TEST=make buildall, manually check discarded symbols in the .map file
and run on Spring and Link.
The size delta is the following:
Link: total 85.7k -> 84.9k (.text 60.3k -> 59.5k)
Spring: total 59.2k -> 57.2k (.text 44.4k -> 42.5k)
Change-Id: Ib6eb0d3f2cc4fc172c9fc26acac2e486921690a3
Reviewed-on: https://chromium-review.googlesource.com/189224
Reviewed-by: Randall Spangler <rspangler@chromium.org>
Commit-Queue: Vincent Palatin <vpalatin@chromium.org>
Tested-by: Vincent Palatin <vpalatin@chromium.org>
